China's urban air quality evaluation with streaming data: A DEA window analysis.

Area of Science:

• Environmental Science
• Urban Development
• Data Science

Background:

• Rapid urbanization in China has led to significant air pollution challenges.
• Environmental detection systems generate vast amounts of real-time air quality data.
• Existing methods struggle with the heterogeneity of time-varying air quality data.

Purpose of the Study:

• To develop a dynamic air quality index for Chinese cities.
• To analyze the spatiotemporal patterns of urban air pollution in China.
• To identify factors influencing air quality variations.

Main Methods:

• Constructed a window Data Envelopment Analysis (DEA) model.
• Applied hierarchy analysis to handle time-varying data heterogeneity.
• Utilized daily air quality data (CO, NO, SO2, PM2.5, PM10) for 360 Chinese cities from January 2018 to August 2019.

Main Results:

• Over 95% of analyzed cities exceeded air pollution emission standards by at least 30%.
• Urban air quality in China is significantly influenced by monthly variations.
• An inverse U-shaped relationship exists between annual air quality and time.
• Air quality is not significantly affected by the order of weeks within a month or days within a week.
• Pollution hotspots are concentrated in central China, forming a continuous zone centered in Shanxi province.

Conclusions:

• Urban air pollution remains a critical issue in most Chinese cities.
• The developed window DEA model effectively computes dynamic air quality indices.
• Understanding temporal patterns (monthly, yearly) is crucial for air quality management.
• Geographical concentration of pollution in central China requires targeted interventions.